Exploratory assessment of the chemical and antioxidant potential of Apis mellifera honeys from Santa Cruz, Bolivia

Abstract

This preliminary exploratory study investigates the chemical composition and antioxidant potential of Apis mellifera honeys from Santa Cruz, Bolivia, a region with notable honey production but limited data. Eleven samples collected between 2023 and 2024 were processed to obtain enriched extracts (EE) using solid-phase extraction. Chemical profiles were obtained by TLC and HPLC. Total phenolics, flavonoids, and antioxidant capacity (DPPH•, TEAC, FRAP) were determined spectrophotometrically. Profiles revealed compositional differences, with tentative identification of phenolic acids, hydroxycinnamic acids, and flavonoid derivatives. Total phenolics ranged from 3.57 to 24.95 mg GAE/100 g of honey, and TEAC from 0.37 to 2.10 μmol TE/g honey. Notably, sample M11, with Tessaria spp. reported as a dominant floral source, exhibited the highest antioxidant potential, suggesting interest for functional applications. Darker honeys generally had higher antioxidant capacity, though not always reflecting chemical diversity. These findings highlight the complex bioactive composition of Bolivian honeys and the role of floral and environmental factors.

Graphical abstract

Highlights

• •Amberlite resin SPE enhanced polyphenol recovery from Bolivian honey samples.
• •Phenolic acids and flavonoid profiles varied across Santa Cruz honeys.
• •TPC, TFC, and antioxidant capacity differed notably among samples.
• •First chemical–antioxidant report on Bolivian honeys, baseline for future studies.

1. Introduction

Honey is primarily composed of carbohydrates, accompanied by smaller quantities of organic acids, amino acids, minerals, vitamins, volatile compounds, and phytochemicals, including bioactive compounds (de Almeida-Muradian et al., 2020). These naturally occurring substances can modulate metabolic processes and contribute to health benefits upon consumption (Dincheva et al., 2023). Their characteristics and biological activities, such as antioxidant, anti-inflammatory, antimicrobial, and antitumor effects, depend on compound type (e.g., polyphenols, carotenoids, tocopherols, phytosterols) and chemical structure (Drețcanu et al., 2022).

Among the bioactive compounds found in Apis mellifera honeys, a variety of phenolic acids (e.g., salicylic, caffeic, p-coumaric, and vanillic) and flavonoids (e.g., quercetin, rutin, and naringenin) have been reported (Lawag et al., 2022; Missio da Silva et al., 2020; Viteri et al., 2021). The botanical and geographical origins of honey are key factors shaping its physicochemical profile and antioxidant potential. For instance, honeys derived from eucalyptus and mastic tree sources exhibit higher phenolic content and antioxidant activity (do Nascimento et al., 2018), while darker honeys generally show greater concentrations of bioactive compounds than lighter varieties (Galhardo et al., 2021). Furthermore, honeys of similar floral origin may vary significantly in chemical composition and antioxidant performance due to environmental and geographical differences (Mohammed, 2020). However, many studies addressing this variability rely primarily on global antioxidant parameters, without incorporating chromatographic profiling to explore differences in phenolic composition among samples.

According to the Food and Agriculture Organization (FAO), Bolivia ranks 84th globally in honey production, with an output of approximately 701.32 t in 2023 (FAO, 2023). Although several studies have examined the organoleptic, physicochemical, and antioxidant properties of honeys from different regions of the country (Echalar Baldiviezo, 2014; Quino & Alvarado, 2017), the specific identification and characterization of the bioactive compounds contributing to these characteristics, particularly through chromatographic approaches, remain limited, specifically in samples from Santa Cruz, a department known for its notable honey production and biodiversity (Ministerio de Educación, 2015).

Bolivia's rich ecological diversity is reflected in its classification into multiple ecoregions, distinct ecological areas defined by unique combinations of vegetation, climate, soil, and fauna (Ibisch et al., 2003). Santa Cruz alone encompasses eight such ecoregions, which may influence the availability of floral resources and, consequently, the chemical composition of local honeys. Studying honeys from these diverse areas may therefore enable the discovery of unique bioactive compounds with potential applications in pharmaceutical (Lomartire & Gonçalves, 2022), food (Zhou et al., 2022), or cosmetic formulations (Alharbi et al., 2021), while providing added value to local apiculture.

In addition to their scientific and functional relevance, identifying distinctive chemical profiles in honeys from Santa Cruz may enhance their marketability by emphasizing their bioactive and antioxidant potential. Such differentiation could facilitate access to value-added markets, ultimately supporting the sustainable development of the apicultural sector and contributing to the economic empowerment of small-scale beekeepers in the region. Despite the growing interest in honey phenolics worldwide, Bolivian honeys, particularly those from Santa Cruz, remain underexplored in terms of their detailed chemical and antioxidant profiles. To address this gap, the present study adopts an integrated exploratory approach combining solid-phase extraction (SPE) using Amberlite XAD resin, chromatographic profiling (TLC and HPLC-DAD), and multiple antioxidant assays. Specifically, this study aims to explore the variability in antioxidant properties and bioactive compound content of A. mellifera honeys from different localities in Santa Cruz. We hypothesize that these properties differ significantly across sampling sites, reflecting environmental heterogeneity among ecoregions.

2. Materials and methods

2.1. Sample collection and preliminary analysis

Eleven A. mellifera honey samples were provided by local beekeepers engaged in commercial production from various localities across the department of Santa Cruz, Bolivia, spanning seven distinct ecoregions (Fig. 1). Samples were harvested between March 2023 and March 2024. Local beekeepers provided supplementary data, including the geographic coordinates of apiaries, predominant floral sources, and harvest dates. Geographic verification of ecoregion classification was conducted using QGIS software. While this set of samples does not fully represent the total honey production of Santa Cruz, it provides a preliminary exploratory dataset covering a diversity of ecoregions and floral sources.

Fig. 1.

Location of apiaries corresponding to 11 honey samples of A. mellifera provided by local beekeepers from the department of Santa Cruz, Bolivia. Modified from Ibisch et al. (2003).

Two hundred grams of each honey sample was stored in sterile, sealed containers under dark conditions at 6–8 °C until analysis. Physicochemical parameters, including moisture content, pH, and color, were assessed. Moisture was determined gravimetrically after drying at 105 °C for 24 h (AOAC, 2000). pH was measured after dilution in distilled water using a pH meter (Galhardo et al., 2021). Honey color was measured spectrophotometrically at 635 nm and expressed as mm Pfund (Bodor et al., 2021).

Enriched extracts (EEs) were obtained through solid-phase extraction following a previously described protocol (Martos et al., 2000). 50 g of each honey sample was diluted in 250 mL of acidified water (pH 2) and filtered to remove particulates. Filtrates were mixed with 100 g of pre-activated Amberlite XAD-7 resin (activated with acidified water for 30 min) and stirred for 20 min to adsorb phenolic compounds while sugars and other interferences were removed with aqueous washes. Phenolics were then desorbed with three washes of 200 mL of ethanol, and the eluates were concentrated under reduced pressure using a rotary evaporator. The Amberlite XAD-7 resin was reused for subsequent samples after thorough washing with alkaline and acid solutions, and reactivation with acidified water. Extracts were stored at −80 °C until lyophilization. Extraction yields were recorded, and the EEs were subsequently used for chemical profiling and antioxidant analysis.

2.2. Chemical profile of polyphenols

Thin-layer chromatography (TLC) was used for the preliminary separation and qualitative identification of bioactive compounds in the enriched extracts (EEs). Pre-coated silica gel 60 F₂₅₄ TLC plates (0.20 mm layer thickness; Macherey-Nagel, Germany) were used as the stationary phase, cut to a uniform height of 5 cm, with variable width depending on the number of samples applied. EEs were prepared at 5 mg/mL in a 1:1 (v/v) water:ethanol solution. Various organic solvent systems were tested as mobile phases depending on sample polarity. TLC plates were developed in an ascending mode in a saturated TLC chamber until the solvent front reached 0.5 cm from the top of the plate (4 cm migration distance) and were visualized under UV light at 254 and 366 nm. Plates were subsequently treated with chemical reagents to detect specific compound classes: 2-aminoethyl diphenylborinate (NPR reagent) for phenolic compounds and p-anisaldehyde for other compounds such as terpenes (Wagner & Bladt, 1997). Retention factors (Rf) were compared among samples and with known standards (trans-cinnamic acid, caffeic acid, catechin, quercetin, quercetin glucuronide, rutin, kaempferol and kaempferol 3-O-glucoside (Fig. S1), which were prepared at 100 μg/mL.

High-performance liquid chromatography (HPLC) coupled to a diode array detector (Azura, Knauer, Germany) was used for further separation and characterization of bioactive compounds in the EEs. The method was adapted from Machado De-Melo et al. (2018), using a Kromasil® C18 reversed-phase column 100 Å (5 μm, 4.6 mm × 150 mm), operated at 30 °C. The mobile phases consisted of 0.1% formic acid in water (A) and 0.1% formic acid in a 1:1 methanol:acetonitrile mixture (v/v) (B). The elution was performed using a gradient program as follows: starting at 15% B, increasing to 30% B at 20 min, 45% B at 40 min, 50% B at 45 min, 55% B at 50 min, 70% B at 65 min, reaching 90% B between 75 and 80 min, and returning to 15% B at 85 min. Extracts were prepared at 5 mg/mL in methanol:water (1:1), and 20 μL were injected at a flow rate of 0.5 mL/min. Chromatograms were monitored at 270, 330, and 360 nm, and UV–Vis spectra (200–600 nm) were recorded for signal characterization. No authentic standards were used for compounds identification. Peak assignment was therefore tentative and based exclusively on retention behavior and comparison of UV–Vis spectral features with characteristic phenolic profiles reported in the literature, without confirmation by mass spectrometry (MS or HRMS). This approach was consistent with exploratory characterization purposes.

2.3. Total phenolic and flavonoid content

The total phenolic content (TPC) of the samples was quantified using the Folin–Ciocalteu spectrophotometric method, following the microassay procedure described by Rizvi et al. (2023) with minor modifications. Gallic acid was used for calibration (1–100 μg/mL, R² = 0.9953), and absorbance was measured at 765 nm. Extracts were prepared at a concentration of 5 mg/mL in ethanol:water (1:1). The reaction mixture was prepared in a 1:1:8 proportion of extract, Folin–Ciocalteu reagent (2 N), and sodium carbonate solution (5% w/v), respectively. The mixture was incubated at 37 °C in the dark for 30 min, and absorbance was read using a microplate spectrophotometer (Biotek Epoch, Agilent Technologies, CA, USA). Results were expressed as mg of gallic acid equivalents (GAE) per 100 g of honey.

Total flavonoid content (TFC) was assessed using the aluminum chloride (AlCl₃) colorimetric method as described by Shraim et al. (2021) with modifications for microplate use. A quercetin calibration curve (5–70 μg/mL, R² = 0.9951) was constructed, and absorbance was measured at 430 nm. The reaction mixture was prepared using a 1:1 proportion of extract (5 mg/mL in ethanol:water 1:1) and AlCl₃ solution (2% w/v). After a 60-min incubation in the dark, absorbance was measured at the same wavelength. Results were expressed as mg of quercetin equivalents (QE) per 100 g of honey.

2.4. Antioxidant activity

To evaluate the antioxidant capacity of honey extracts, spectrophotometric assays (Biochrom Libra S60PC, Biochrom Ltd., Cambridge, United Kingdom) based on different antioxidant mechanisms were conducted: DPPH^•, TEAC, and FRAP.

The DPPH^• (2,2-Diphenyl-1-picrylhydrazyl) assay was performed according to Larsen and Ahmed (2022) with slight modifications. A 20 mg/L DPPH^• solution in ethanol was mixed with enriched extracts (EE) at 300 μg/mL in a 2:1 ratio. After a 5-min incubation at room temperature, absorbance was measured at 517 nm. Results were expressed as the percentage of DPPH^• radical scavenging.

The TEAC (Trolox Equivalent Antioxidant Capacity) assay, based on ABTS^•+ decolorization, followed the method by Rumpf et al. (2023). A Trolox calibration curve (100–2000 μM, R² = 0.9988) was prepared and absorbance was measured at 734 nm. ABTS^•+ was generated by mixing 5 mL of ABTS (7 mM) with 88 μL of potassium persulfate (140 mM), incubated for 16 h in the dark at room temperature. This solution was diluted with ethanol to an absorbance of 0.70 ± 0.05. EE solutions at different concentrations (125–1000 μg/mL) were mixed with ABTS^•+ in a 1:100 proportion and incubated in the dark for 6 min. Absorbance was measured at 734 nm, and results were expressed as μmol Trolox equivalents (TE) per gram of honey.

The FRAP (Ferric Reducing Antioxidant Power) assay followed Rao et al. (2023) with some adjustments. A calibration curve using Trolox (14–295 μM, R² = 0.9927) was constructed at 593 nm. The FRAP reagent was prepared by mixing acetate buffer (300 mM), TPTZ solution (40 mM in HCl), and FeCl₃·6H₂O (20 mM) in a 25:5:10 ratio (v/v/v). EE at 700 μg/mL was mixed with the FRAP solution in a 1:19 proportion, incubated in the dark for 30 min, and the absorbance was recorded at 593 nm. Results were also expressed as μmol TE per gram of honey.

2.5. Statistical analysis

Physicochemical parameters were determined in duplicate, while all other assays were performed in triplicate. Rf values of TLC chromatograms were calculated using ImageJ software. Results were expressed as means ± standard deviation. Statistical analyses were conducted using GraphPad Prism 10. A one-way ANOVA followed by Tukey's post-hoc test was used to evaluate significant differences (p < 0.05) among honey samples in terms of TPC, TFC, and antioxidant capacity. Pearson correlation coefficients were also calculated to assess the relationships between antioxidant capacity, TPC content, and selected physicochemical parameters.

3. Results and discussions

3.1. Sample collection and preliminary analysis

For this preliminary exploratory study and given the diversity of ecoregions and floral sources in Santa Cruz, Table 1 summarizes detailed information on the analyzed A. mellifera honey samples, including geographic coordinates, local (vernacular) vegetation with tentative Latin names (Genus spp.), harvest period, and preliminary analyses, including moisture content, pH, and color (Pfund scale). All samples correspond to floral honeys produced by A. mellifera. The botanical origin (monofloral vs. multifloral) could not be confirmed, as no melissopalynological analysis was performed; therefore, vegetation information is limited to the predominant local flora reported by beekeepers in the vicinity of the apiaries and is provided solely for contextual purposes.

Table 1.

Characteristics of the eleven honey samples collected in Santa Cruz, Bolivia.

Sample	Longitude	Latitude	Ecoregion	Local vegetation	Latin name (unconfirmed)	Honey harvest	pH	Moisture (%)	Color (mm)	Pfund classification
M1	−63°31′05.6”	−17°52′51.1”	Bosque Seco Chiquitano	Cari Cari Curupaú Cítricos	Machaerium spp. Anadenanthera spp. Citrus spp.	INA	4,0 ± 0,0F	18,9 ± 0,3C	6,7 ± 1,2E	Water white
M2	−62°34′04.1”	−16°26′33.8”	Bosque Seco Chiquitano	Cuchi Tinto Turere Picana Negra	Astronium spp. Callisthene spp. Rhamnidium spp. Cordia spp.	Spring 2023	4,4 ± 0,0D	17,1 ± 0,1DE	17,1 ± 1,1D	White
M3	−63°02′05,0”	−16°55′53.3”	Bosque Seco Chiquitano	Cuchi Paichané Algarrobo	Astronium spp. Vernonanthura spp. Prosopis spp.	Spring 2023	3,8 ± 0,0G	19,6 ± 0,3BC	23,4 ± 3,2C	White
M4	−62°51′81.1”	−16°26′51.2”	Cerrado	Cuchi	Astronium spp.	Spring 2023	4,6 ± 0,0B	17,4 ± 0,1D	44,1 ± 0,7B	Extra light-amber
M5	−63°43′88.8”	−18°46′30.3”	Gran Chaco	Cuchi	Astronium spp.	Summer 2024	4,0 ± 0,0F	20,5 ± 0,1AB	17,0 ± 2,3D	White
M6	−62°82′16.5”	−17°63′58.0”	Gran Chaco	Parajobobo Cupesi Coco Espino Blanco	Tessaria spp. Prosopis spp. Guazuma spp. Acacia spp.	Spring 2023	4,5 ± 0,0C	19,1 ± 0,4C	26,8 ± 3,6C	White
M7	−62°82′16.5”	−17°63′58.0”	Gran Chaco	Parajobobo Cupesi Coco Espino Blanco	Tessaria spp. Prosopis spp. Guazuma spp. Acacia spp.	Summer/Autumn 2024	3,7 ± 0,0H	16,1 ± 0,3E	−11,8 ± 0,9G	Water white
M8	−63°28′08.0”	−20°23′33.8”	Chaco Serrano	Lecherón	Sapium spp.	Autumn 2024	5,6 ± 0,0 A	19,3 ± 0,1C	45,0 ± 1,1B	Extra light-amber
M9	−63°63′47.1”	−18°10′93.6”	Yungas	Cuchi Curupaú Paichané	Astronium spp. Anadenanthera spp. Vernonanthura spp.	Summer/Autumn 2024	3,9 ± 0,0F	17,2 ± 0,1D	−0,7 ± 1,7F	Water white
M10	−63°84′33.9”	−18°21′37,4”	Bosque Tucumano Boliviano	Tipa Soto Curo-Curo Villca Paichané	Machaerium spp. Schinopsis spp. INA Anadenanthera spp. Vernonanthura spp.	Spring 2023	4,3 ± 0,0D	19,7 ± 0,6AC	−8,1 ± 0,9G	Water white
M11	−63°89′05.1”	−16°12′68.2”	Sabanas Inundables	Parajobobo	Tessaria spp.	Autumn 2023	4,1 ± 0,0E	20,7 ± 0,2 A	105,5 ± 3,4 A	Amber

INA = Information not available. Different letters (A, B, C, D, E, F, G, H) in a column denote statistically significant differences among the samples, as determined by Tukey's post-hoc test (p < 0.05).

Physicochemical analysis showed that pH values ranged from 3.70 to 5.62, consistent with previously reported acidity levels for honey. Sample M8 exhibited the highest pH value, suggesting a lower concentration of organic acids or a higher mineral content. Although uncommon, honeys with pH values up to 6.5 have been reported and remain microbiologically active in inhibiting microbial growth (Machado De-Melo et al., 2018).

Moisture content varied between 16.10% and 20.72%, meeting or closely approaching the maximum moisture limits set by the Codex Alimentarius Standard for Honey (Codex Alimentarius Commission, 1981). Environmental factors such as ambient humidity, high temperatures, and storage conditions likely influenced these values (Mohammed, 2020). Furthermore, enzymatic oxidation of glucose by glucose oxidase, producing gluconic acid and hydrogen peroxide, may have contributed to slight changes in water content over time (Missio da Silva et al., 2020).

Color analysis using the Pfund scale revealed values ranging from −11.84 mm (water white) to 105.52 mm (amber). Color variability among samples may reflect differences in floral origin, polyphenolic and carotenoid content, and moisture levels (Bodor et al., 2021). Darker honeys have frequently been associated with higher phenolic and flavonoid concentrations, as well as greater antioxidant capacity. Moreover, color is a sensory attribute that strongly influences consumer preference, with lighter honeys favored in some markets and darker varieties in others (Becerril-Sánchez et al., 2021).

EEs were successfully obtained from all 11 honey samples. The extraction yield was calculated based on the weight of the lyophilized extracts relative to the 50 g of honey used for each sample. The extraction yields ranged from 0.22% to 1.59%, with an average of 0.71%. These variations may be attributed to differences in the chemical composition of the honeys. Furthermore, the generally low yields (<2%) reflect the high sugar content of honey, with phytochemicals present in much smaller quantities (de Almeida-Muradian et al., 2020).

The use of Amberlite resin as a solid-phase extraction method is widely reported for isolating polyphenols from diverse matrices. This method can recover up to 90% of flavonoids while efficiently removing sugars, acids, pigments, and other compounds, in contrast to other extraction techniques. Moreover, the use of acidified water during extraction minimizes phenolic oxidation, enhancing the recovery of polyphenolic compounds (Pascual-Maté et al., 2018). Given the low concentration of polyphenols in honey, detecting these compounds would have been more challenging without the enrichment process.

Recently, biodegradable resins have emerged as a promising alternative, offering higher extraction efficiency and stronger antioxidant activity in the resulting extracts compared to Amberlite. In addition to their performance, these biosorbents offer a more sustainable approach for phenolic recovery in honey and similar complex matrices (Neggad et al., 2021).

3.2. Chemical profile of polyphenolic compounds

To investigate compositional differences potentially underlying the variation in antioxidant properties, EEs were analyzed using TLC. Among the solvent systems tested, mobile phase A (ethyl acetate:acetic acid:water, 10:1.2:1) and mobile phase B (chloroform:ethyl acetate:methanol, 7:1:2) provided the best resolution (Fig. 2).

Fig. 2.

TLC profiles of the EEs with different physical and chemical developers, and the corresponding Rf values. Mobile phases: A (ethyl acetate:acetic acid:water, 10:1.2:1) and B (chloroform:ethyl acetate:methanol, 7:1:2).

Qualitative differences in chemical composition were observed both among samples from the same ecoregion and between different ecoregions. Some samples displayed a greater number of spots, suggesting higher chemical diversity. Using mobile phase B and shortwave UV light (254 nm), spots corresponding to non-fluorescent compounds appeared at Rf of 0.81 and 0.77 in samples M2, M4, M5, M8, and M10. After revealing with p-anisaldehyde, similar blue-violet spots appeared, possibly indicating the presence of terpenoid compounds (Santiago & Strobel, 2013).

Under longwave UV light (366 nm), fluorescent spots were detected, and their intensity slightly increased after spraying with NPR reagent. Similar fluorescence patterns were observed between samples M2 and M4, and between M6, M7, and M8. Comparison with authentic standards run under the same conditions showed that, although Rf values were not coincident, the fluorescence color and behavior after NPR derivatization were comparable. In particular, blue fluorescence was associated with hydroxycinnamic acid derivatives, whereas green/yellow fluorescence suggested the presence of flavonol derivatives. Therefore, compound assignment was based on qualitative fluorescence characteristics rather than exact Rf matching. No spots were revealed in samples M3 and M9, possibly due to compounds absorbing at different wavelengths or lacking UV absorbance.

To tentatively identify bioactive compounds, HPLC–DAD resolved a total of 28 chromatographic peaks across all EE (Table 2). These peaks were classified based on their retention times and UV–Vis spectral features. Samples M2, M4, and M8 exhibited the highest number of peaks, indicating greater chemical complexity, while M3, M7, and M9 showed the fewest. Although most analytes exhibited maximum absorbance at 270 nm (Fig. 3), extracts M5 and M10 presented additional or more intense peaks at 330 nm and 360 nm (Fig. S2), features consistent with hydroxycinnamic acids and flavonoids (Mabry et al., 1970). In contrast, these longer-wavelength peaks were absent in other samples, consistent with the lack of fluorescence in their TLC profiles.

Table 2.

Preliminary identification of bioactive compounds in A. mellifera honeys.

Signal	Rt(min)	UVmax(nm)	M1	M2	M3	M4	M5	M6	M7	M8	M9	M10	M11	Preliminary identification
1	6,7	264			x									Hydroxybenzoic acid derivative
2	7	284	x	x		x	x	x	x		x	x		Hydroxybenzoic acid derivative
3	10,8	279		x										Flavanol derivative
4	10,8	330					x					x		Not identified
5	11,3	277				x								Flavanol derivative
6	11,7	330	x											Not identified
7	12,3	330					x					x		Not identified
8	14,4	284		x		x				x				Hydroxybenzoic acid derivative
9	14,5	254					x							Hydroxybenzoic acid derivative
10	14,9	284				x								Hydroxybenzoic acid derivative
11	16,4	274		x		x				x				Flavanol derivative
12	16,4	300sh, 310					x						x	Hydroxycinnamic acid derivative
13	21	274								x				Not identified
14	23,2	275, 310								x				Flavanone derivative
15	23,9	260, 374					x							Flavonol derivative
16	24,6	284sh, 324	x											Hydroxycinnamic acid derivative
17	24,9	275								x				Hydroxybenzoic acid derivative
18	26,2	274											x	Not identified
19	27,8	254		x										Not identified
20	28,8	254, 299		x										Flavone derivative
21	29,5	254, 299				x								Flavone derivative
22	31,4	274											x	Not identified
23	33,1	284sh, 330					x							Hydroxycinnamic acid derivative
24	33,8	274		x						x				Not identified
25	34,4	254		x						x				Not identified
26	34,5	274				x								Not identified
27	35,3	254				x								Not identified
28	37,8	284sh, 330					x							Hydroxycinnamic acid derivative

Fig. 3.

HPLC chromatographic profiles monitored at 270 nm.

By comparing retention times and UV–Vis spectra, compounds were tentatively assigned to six classes: hydroxybenzoic acid derivatives, hydroxycinnamic acid derivatives, flavanols, flavonols, flavanones, and flavones (Mabry et al., 1970). The analyses were performed on a reverse-phase column, allowing the separation of compounds according to polarity, with the most polar eluting earlier and the less polar retained longer. Hydroxybenzoic derivatives were found in all samples except M11, while hydroxycinnamic derivatives appeared only in M1, M5, and M8. Flavonoid diversity was limited: flavanol and flavonol peaks were detected exclusively in M2, M4, M5, and M8, and flavanones and flavones were even less frequently observed. A few peaks did not match any of these groups, suggesting the possible presence of other bioactive families.

Overall, samples M2, M4, and M8 exhibited similar chromatographic fingerprints despite originating from different ecoregions, suggesting that floral origin alone does not fully determine honey chemistry. Conversely, samples collected from the same reported vegetation sometimes displayed distinct profiles, likely influenced by minor floral contributors or differences in harvest timing. However, the limited phytochemical characterization of the flora in Santa Cruz hinders the identification of specific compounds and meaningful comparisons with honeys from other regions.

Despite this limitation, the consistent presence of certain peaks in specific samples suggests their potential as botanical or geographical markers. Similar approaches have been employed elsewhere to authenticate monofloral honeys. For instance, chlorogenic acid, as well as gallic acid and protocatechuic acid, can serve as chemical markers for Greek chestnut honeys and oak-derived honeydew honeys, respectively (Ntakoulas et al., 2025). In Manuka honey (Leptospermum scoparium), leptosperin and leptosin have been reported exclusively in samples from Oceania (Viteri et al., 2021). Likewise, kaempferitrin was exclusively identified in Camellia oleifera honey and not detected in nine other monofloral honeys, supporting its role as a specific flavonoid marker for this floral source (Li et al., 2023).

In contrast, honeys from Santa Cruz have not yet been analyzed for specific flavonoids, limiting the development of regional botanical markers. Nonetheless, the exploratory results of this study suggest that some samples exhibit distinctive chemical profiles that merit further investigation. Future research, including targeted high-resolution mass spectrometry and monofloral honey analysis, is necessary to validate candidate compounds and support the traceability and quality assurance of Bolivian honeys.

3.3. Total phenolic and flavonoid content

Table 3 summarizes the quantitative results obtained for TPC, TFC, and antioxidant activity (DPPH^•, TEAC/ABTS, and FRAP assays) in the EEs analyzed. Detailed correlation analyses are shown in Fig. S3.

Table 3.

TPC, TFC and antioxidant activity of EE measured by DPPH, TEAC and FRAP assays.

Sample	TPC (mg GAE/100 g honey)	TFC (mg QE/100 g honey)	DPPH (% inhibition)	TEAC (μmol TE/g honey)	FRAP (μmol TE/g honey)
M1	7.5 ± 0.4F	1.25 ± 0.03G	14.5 ± 1.1E	0.37 ± 0.03E	0.43 ± 0.04FG
M2	7.6 ± 0.3F	1.28 ± 0.03G	33.4 ± 0.7AB	0.57 ± 0.05E	0.52 ± 0.05EF
M3	16.3 ± 1.0BD	3.0 ± 0.2C	20.2 ± 1.4D	1.2 ± 0.1B	0.73 ± 0.07DE
M4	17.2 ± 0.7B	3.44 ± 0.05B	35.3 ± 1.4 A	1.08 ± 0.05BC	0.95 ± 0.08CD
M5	10.5 ± 0.3E	2.04 ± 0.06EF	24.7 ± 1.6CD	0,9 ± 0,1CD	0.89 ± 0.08CD
M6	14.7 ± 0.9CD	3.7 ± 0.2B	25.3 ± 2.3C	0.89 ± 0.06CD	1.1 ± 0.1C
M7	3.6 ± 0.3G	0.79 ± 0.03H	30.2 ± 3.1B	0.40 ± 0.04E	0.26 ± 0.02G
M8	16.8 ± 0.1BC	2.34 ± 0.01DE	34.0 ± 2.0AB	1.16 ± 0.06B	1.42 ± 0.08B
M9	8.1 ± 0.5F	1.73 ± 0.07F	8.1 ± 0.7F	0.43 ± 0.04E	0.9 ± 0.1CD
M10	14.2 ± 1.4D	2.57 ± 0.09D	0.0 ± 0.0G	0.8 ± 0.2D	1.42 ± 0.08B
M11	25.0 ± 1.2 A	7.7 ± 0.4 A	36.3 ± 3.6 A	2.1 ± 0.1 A	2.6 ± 0.2 A

Different letters (A, B, C, D, E, F, G, H) denote statistically significant differences among the samples, as determined by Tukey's post-hoc test (p < 0.05).

The TPC of the EEs ranged from 3.57 to 24.95 mg GAE/100 g of honey, while the TFC ranged from 0.79 to 7.72 mg QE/100 g of honey. Sample M7 exhibited the lowest TPC and TFC values, whereas M11 showed the highest for both. A strong positive correlation was observed between TPC and TFC (R² = 0.8172), suggesting that a substantial portion of phenolic compounds may be flavonoids, despite the low chemical diversity observed in most samples based on chromatographic profiles. Additionally, a correlation between TFC and Pfund values (R² = 0.7888) indicates that darker honeys tended to have higher TFC concentrations.

Variations in TPC and TFC were observed not only between ecoregions but also among samples from the same ecoregion. For instance, although M1 and M2 had similar TPC values, their chemical profiles differed significantly, and neither resembled M3 from the same ecoregion. Samples M6 and M7, collected from the same apiary in different seasons, also differed: the spring sample (M6) showed higher TPC than the autumn one (M7). The darkest samples (M4, M8, and M11) also exhibited the highest TPC values.

Interestingly, TPC and TFC values did not always correlate with chemical profile complexity. M7, with the lowest concentrations, only contained a hydroxybenzoic acid derivative and no flavonoids were identified. M2 had similarly low TPC and TFC but a complex chemical profile, while M11 had the highest concentrations but limited chemical diversity. These results suggest that both compound type and concentration may influence bioactive properties.

These differences may reflect the foraging behavior of A. mellifera, which can fly 2 km from the hive depending on climatic conditions, wind, temperature, and floral resource availability (Vincze et al., 2025). Thus, even within a single ecoregion, variability in floral diversity and flowering periods can influence honey composition.

No significant correlations were found between TPC and moisture or pH. Previous studies have shown that honey moisture tends to increase over storage due to hygroscopic properties, while pH changes slightly yet phenolic content remains stable, indicating no direct relationship with moisture content (Missio da Silva et al., 2020).

Compared to Brazilian honeys, which is close to Santa Cruz, the TPC values observed here were lower, while TFC values were similar or slightly lower. In Rio Grande do Sul State, TPC ranged from 26.0 to 100 mg GAE/100 g honey (do Nascimento et al., 2018), and in Paraná State from 143.67 to 191.17 mg GAE/100 g (Galhardo et al., 2021). TFC values ranged from 0.65 to 8.10 mg QE/100 g and 7.97–44.99 mg QE/100 g, respectively. In both cases, darker honeys had higher polyphenol content, and harvest season influenced outcomes. The comparison with Brazilian honeys is especially relevant due to the geographical proximity and shared ecological features, such as similar vegetation types and ecosystems with Santa Cruz.

Since these compounds originate from plants, differences in TPC and TFC likely stem from botanical variation near the hives. A. mellifera may collect nectar from one or multiple plant species, with content also influenced by environmental conditions (Zaldivar-Ortega et al., 2024, do Nascimento et al., 2018). For instance, Anadenanthera spp. (Curupaú) contains catechin, quercetin, and kaempferol (Maia et al., 2024), while Tessaria spp. (Parajobobo) contains flavonoids such as quercetin and naringenin and phenolic acids like gallic and vanillic acid (Sosa-Lochedino et al., 2022), though concentrations vary by plant organ.

Regarding methodology, many studies use aqueous dilutions without extract preparation (Galhardo et al., 2021; Missio da Silva et al., 2020). In such cases, reducing sugars may react with the Folin–Ciocalteu reagent, producing blue complexes and leading to false positives for TPC (Munteanu & Apetrei, 2021). Here, solid-phase extraction was employed to eliminate these interferences. Similarly, improper use of the AlCl₃ colorimetric assay for TFC can lead to inaccuracies, as the formation of complexes depends on the structural properties of both the reference standard and the flavonoids present (Shraim et al., 2021).

3.4. Antioxidant activity

The antioxidant capacity of the enriched honey extracts varied according to the chemical mechanism targeted by each assay. At a concentration of 300 μg/mL, none of the samples achieved 50% inhibition of the DPPH^•. Sample M10 showed no inhibition, while M9 exhibited the second-lowest activity. In contrast, samples M2, M4, M8, and M11 displayed the highest antioxidant activity, with inhibition percentages ranging from 33% to 36%. Except for M2, these samples were characterized by darker coloration.

TEAC assay results ranged from 0.37 to 2.10 μmol TE/g honey. The lowest activities were observed in M1, M2, M7, and M9, while M11 demonstrated the highest ABTS^•+ scavenging capacity. Similarly, FRAP values were lower overall (0.08–0.91 μmol TE/g), with M1, M2, and M7 again among the least active, and M11 showing the greatest ferric reducing ability.

No correlation was observed between total phenolic content (TPC) and DPPH^• activity. However, strong positive correlations were found between TPC and both TEAC (R² = 0.8996) and FRAP (R² = 0.7369). A direct correlation also existed between TEAC and FRAP values, whereas DPPH^• showed no significant relationship with either assay.

These discrepancies reflect the differing chemical sensitivities of the assays. DPPH^• is preferentially reduced by lipophilic, nonpolar antioxidants such as ferulic, sinapic, and p-coumaric acids, or naringenin. In contrast, TEAC is responsive to both hydrophilic and lipophilic compounds, while FRAP primarily detects hydrophilic antioxidants like gallic acid, caffeic acid, quercetin, and kaempferol. Additional factors, such as glycoside-bound compounds and assay pH, may also influence antioxidant activity (Munteanu & Apetrei, 2021).

Variability in antioxidant capacity was observed both within and between ecoregions. For instance, sample M2 exhibited low TPC but relatively high DPPH^• inhibition, while M10 showed no DPPH^• activity despite moderate performance in TEAC and FRAP assays. These findings suggest that beyond phenolic concentration, the chemical diversity of bioactive constituents, likely shaped by surrounding floral sources, plays a crucial role.

Notably, none of the extracts achieved 50% inhibition in the DPPH^• assay at the tested concentration, contrasting with studies of diluted honeys from Brazil. do Nascimento et al., (2018) reported IC₅₀ values ranging from 25.45 to 294.26 mg/mL. Some studies have even documented negative correlations between TPC and DPPH^• activity, indicating that phenolics are not the sole contributors to antioxidant effects; other components, including organic acids, amino acids, and proteins, may also be involved (Becerril-Sánchez et al., 2021).

TEAC values were comparable to other A. mellifera honeys. In Brazil, Galhardo et al. (2021) found an average of 1.0 ± 0.3 μmol TE/g honey, attributing variation to floral origin, geography, and seasonality. Similarly, in Italy, Attanzio et al. (2016) reported averages between 0.19 and 2.70 μmol TE/g honey produced by Sicilian black honeybees, from acacia, almond and eucalyptus, among other floral sources, highlighting the influence of phytochemical complexity and potential synergistic interactions among honey constituents.

FRAP values aligned with monofloral honeys from Brazil, such as 0.99 ± 0.04 μmol TE/g for Pityrocarpa spp. and 1.8 ± 0.2 μmol TE/g for eucalyptus honeys (de Almeida et al., 2016), and with Ecuadorian eucalyptus honeys (0.34–0.79 μmol TE/g) (Valdés-Silverio et al., 2018). These studies highlight the influence of floral origin, polyphenols, carotenoids, and enzymatic components on antioxidant capacity.

Previous studies have demonstrated the value of integrating botanical, physicochemical, and biochemical analyses for honey characterization. For instance, Gezo honey from Salix spp. in Eastern Anatolia was confirmed as a honeydew honey through melissopalynology, with detailed profiling of phenolic compounds, antioxidant capacity, and antimicrobial and prebiotic activities (Alparslan et al., 2025). Nigella sativa monofloral honey was similarly characterized for total phenolics, flavonoids, antioxidant capacity, and antimicrobial properties, identifying major compounds such as ellagic acid and pinocembrin (Kolayli et al., 2023). In Northern Cyprus, multifloral honeys were analyzed for phenolic and flavonoid content, antioxidant activity, and HPLC-PDA profiles, revealing high diversity linked to floral and geographical origin (Uçar et al., 2023). While these studies provided comprehensive biochemical and functional insights for their respective regions, honeys from Santa Cruz, Bolivia, remain largely unexplored. In this context, our study adopted a preliminary exploratory approach, evaluating chemical composition and antioxidant potential across diverse ecoregions and floral sources.

The observed variability among samples underscores the need for further targeted analyses, including melissopalynology and advanced chromatographic characterization, to establish robust botanical and functional markers for Bolivian honeys. Sample M11, with Parajobobo (Tessaria spp.) reported as a dominant floral source, possibly accounting for its superior antioxidant performance. Its high phenolic content and unique chemical profile reinforce its distinctiveness and potential biological value.

Moisture content and pH showed no significant effect on TPC, TFC, or antioxidant capacity. Instead, these properties depend on the concentration and nature of the bioactive compounds (Missio da Silva et al., 2020). Beekeeping practices can influence these values; for instance, heating honey above 60 °C, commonly applied to prevent crystallization, may increase TPC by releasing phenolic compounds from pollen or through the formation of Maillard reaction products (Islam et al., 2021).

The Maillard reaction occurs when reducing sugars interact with amino groups from proteins or nucleic acids, leading to the formation of glycotoxins and phenol-like derivatives. Heat can also degrade or rearrange phytochemicals, resulting in new antioxidant compounds (Sharma et al., 2022). Therefore, thermal processing, as well as extraction and handling methods during harvest, can significantly alter honey's natural antioxidant profile and contribute to observed variability among samples (Romero, 2024).

Finally, although in vitro assays provide useful preliminary insights into antioxidant potential, they do not replicate in vivo conditions. Under physiological environments, enzymes and metabolic pathways modulate antioxidant activity. Nonetheless, these assays remain valuable tools for characterizing antioxidant capacity in controlled settings.

This study focused on the chemical and antioxidant characterization of honey extracts; therefore, standard quality parameters such as proline content, sugar profile, and enzyme activities were not evaluated and should be addressed in future studies aimed at regulatory or commercial classification.

4. Conclusion

This preliminary, exploratory screening study provided initial insights into the variability of bioactive compounds and antioxidant capacity in commercial A. mellifera honeys from Santa Cruz. The results revealed chemical differences between samples from distinct ecoregions, likely influenced by local flora, harvest season, and beekeeping practices.

Chemical profiling by TLC and HPLC indicated a higher diversity of phenolic acids compared to flavonoid derivatives. Total phenolic content (TPC) and total flavonoid content (TFC) varied among samples and were positively associated with honey color intensity. Antioxidant capacity was low in the DPPH^• assay but moderate in TEAC and FRAP, depending on the chemical nature and concentration of the compounds present.

This study establishes a valuable foundation for future research on bioactive compounds in Bolivian honeys and their potential applications. Of particular interest was sample M11, with Parajobobo (Tessaria spp.) reported as a dominant floral source in Guarayos Province, which exhibited the highest antioxidant compound levels, highlighting the untapped potential of local honeys, especially within the flooded savanna ecoregion. Additionally, spring-harvested honeys from Samaipata (M9, M10) and Montero Hoyos cities (M6, M7) showed higher concentrations of bioactive compounds than autumn samples, underscoring the importance of harvest season and surrounding vegetation in these areas.

To confirm these trends, future research should include a larger number of samples from the same ecoregion and harvest period, integrate melissopalynological analysis to verify floral origin, and combine field data on flowering plants with chemical profiling. This would strengthen the identification of predominant bioactive compounds in each context, while providing a scientific basis for regional valorization, traceability systems, and potential nutraceutical certification of Bolivian honeys.

CRediT authorship contribution statement

Jose A. Limpias-Hurtado: Writing – original draft, Visualization, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Natalia Montellano Duran: Writing – review & editing, Resources, Funding acquisition. Alberto Giménez-Turba: Writing – review & editing, Resources, Funding acquisition. Nélida Nina: Writing – review & editing, Supervision, Project administration, Methodology, Formal analysis, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this manuscript.

Appendix A. Supplementary data

Supplementary material

Data availability

Data will be made available on request.